A novel method for energy storage sizing based on time and frequency domain analysis

Energy storage is used to balance the variant power for the stability of the grid. It is significant to understand the fluctuation characteristic of renewable energy (RE) generation and the requirements of energy storage when large-scale RE is integrated in the grid. In the paper, a novel method based on time and frequency domain analysis is proposed for energy storage system (ESS) sizing, including both power sizing and energy sizing. According to the relationship between charge/discharge power and stored energy, the sizing model is established based on autocorrelation function and power spectral density (PSD) of the stochastic cycling process. The time and spectral characteristic of RE generation is analyzed based on the historical generation data of a wind farm and a PV station in the Northwest region of China. The size of energy storage is determined by the time and frequency domain method respectively. Comparing with the time domain method, it is showed that the frequency domain method is sufficient for energy storage sizing with enough accuracy and a much easier calculation process at the same time.